Oracle licensing: just say no

It's Oracle negotiating season once again and the advisors are out in force. Is what they're saying right or are there better approaches?

In the last week, two reports about Oracle licensing have caught the eye. Both are aimed at the same thing: a mild pitch at negotiation in the run up to Oracle's financial year end. One from Forrester and parsed by Larry Dignan and the other from David Floyer. Both agree on one point although David goes further in providing detail:

Oracle is aggressively pursuing a bundling strategy, and those users interested in that approach can and should negotiate better deals. Despite the obvious lock-in risks, if you're going in this direction Oracle is willing to negotiate with you; especially for large deals over $5M. Virtualization is a strategic initiative for many organizations, and Oracle will be less than supportive. IT organizations need to weigh the risks of Oracle's negative posture toward non-Oracle virtualization strategies with the benefits of virtualizing systems and apps. On balance, for most mainstream applications the benefits outweigh the risks. For mission critical apps the stakes are higher and more caution is warranted.

Larry notes that:

Oracle has to upsell you because it can’t hit its revenue targets just selling new products to willing buyers. There’s a push for more licensing deals. That’s why per-core licensing exists—there’s revenue even if customer rolls shrink.

Forrester seems to take the line that once Oracle has its hooks into you then it is game over for negotiating to any reasonable extent and that the best deals are to be had by pushing forward buying decisions. I fundamentally disagree with that tactic for one simple reason: predicting the future is hard and the risk of acquiring shelfware would require an incredibly compelling business case going down the track. 30% discount now may sound great but when you're paying for something you don't need now is that such a good deal after all?

It has always been the case that for new customers, Oracle would certainly discount on the license but won't play ball on maintenance. It is the license that feeds the annuity stream - or as I prefer to think about it - the entitlement pension. So discounts should not be optional but mandatory. Even so, you have to understand what you're getting into. Unfortunately, it is almost a racing certainty that Oracle's lawyers are better than anything you can field.

But I think there is something else at play here. What Forrester, Larry and John all tacitly acknowledge is that Oracle is primarily driven by its Wall Street masters in order to prop the share price. You can argue that is what all publicly traded companies should be doing. In Oracle's case though, this has led to a highly rigid, centralised and inflexible business model. That in turn means it is vulnerable to attack. David says:

Organizations will find Oracle is much more willing to negotiate if it believes the deal to be competitive. Wikibon knows of several instances where customers have received substantially improved deals by simply initiating (in earnest) IBM DB2 or Microsoft SQL Server pilots and made Oracle aware they are prepared to walk. Caution: We also know of several instances where Oracle has not blinked and customers have been forced to either accept worse terms or migrate major systems off Oracle; a costly endeavor not for the faint of heart.

This is stating the obvious and would be true of any supplier. The problem as I see it is that Oracle has successfully created the myth that once you are even modestly dependent upon Oracle then there is no turning back. DBAs love Oracle because it often represents the perceived full employment act for these folks. Neither of these myths are true. It may initially prove costly to get away from Oracle's clutches but a good business case can usually be made. Whether you want to or need to is a separate discussion. License fees and maintenance are only modest components in an IT budget. However, this discussion would not be on the table if it was not for the pain that some companies feel. The advice here is to undertake a well thought through business needs analysis with correctly costed answers. Your CFO can be your friend here. And since he will usually be a co-signatory, getting him or her in early could reap handsome rewards.

Older versions of Oracle apps need not be maintained by Oracle. If your business is in a moderately stable and steady state then there is no reason why you should not consider a third party maintenance alternative like Rimini Street. Oracle will not like you for it and threaten to penalise you further down the track but do not be intimidated.

Remember that it won't be long before Oracle wants business to switch to Fusion. That's a decision making opportunity that might well leave today's threats looking like empty gestures. It may also leave taking that 30% discount now looking incredibly dumb when you come to review what a Fusion contract contains.

There is another altogether more aggressive approach. Just say no - and while you're at it, enlist the support of your OAUG chair. Do as SAP user groups did and force Oracle to the table rather than blindly putting up with the idea that you have to spend oodles of cash now in order to get a discount. It may not win you a lot in the long term but it will make Oracle understand that for all its market posturing, it doesn't always have to be the only winner.

As a final note, it is not so long since that Oracle positioned itself as a latter day IBM cast in the mold of Tom Watson. Those with long memories know how that ended badly for IBM in the early 1990s. Nobody wishes that of Oracle but right now it is looking dangerously like the old arrogant Big Blue that very nearly fell apart.